I'd like to build a model for german and from what I read up 'til now everyone is talking about acoustic/language model. I'm also a bit confused about the phoneset used - DARPA I guess?
So let's say I want to use SAMPA instead. Is it possible/necessary to build a new phoneset for that reason? Or exists there some mapping between them?
Any help,hint,flame,flower is welcome ;-)
Ralf
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi Ralf.
I have been studying both german and english
at school, and as far as I remember, there is
no exact mapping between english phonemes and
german ones.
So you have two possibilities:
1. you define the best (or the least bad) mapping
from english to german. Yourself.
2. you make your own phone set, based on what
you have in german. For instance, you'll have
to create a phoneme for ch in EIN BUCH., another
one for ch in ICH BIN, and so on for every german
specific sound.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Anonymous
-
2002-06-22
The phoneset used by sphinx is american english. SAMPA should have all the phones in it necessary to do German, but your best bet is to just find an online german dictionary (machine readable) and use it to define the phoneset.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi,
I'd like to build a model for german and from what I read up 'til now everyone is talking about acoustic/language model. I'm also a bit confused about the phoneset used - DARPA I guess?
So let's say I want to use SAMPA instead. Is it possible/necessary to build a new phoneset for that reason? Or exists there some mapping between them?
Any help,hint,flame,flower is welcome ;-)
Ralf
Hi Ralf.
I have been studying both german and english
at school, and as far as I remember, there is
no exact mapping between english phonemes and
german ones.
So you have two possibilities:
1. you define the best (or the least bad) mapping
from english to german. Yourself.
2. you make your own phone set, based on what
you have in german. For instance, you'll have
to create a phoneme for ch in EIN BUCH., another
one for ch in ICH BIN, and so on for every german
specific sound.
The phoneset used by sphinx is american english. SAMPA should have all the phones in it necessary to do German, but your best bet is to just find an online german dictionary (machine readable) and use it to define the phoneset.